Building from knowledge gained and new questions raised during Phases I and II of the Community Treatment Project, Phase III would attempt to systematically determine whether it is feasible and - in terms of improved treatment outcomes - meaningful to (a) broaden the range - and refine the type - of settings and treatment strategies in and through which differential treatment can be carried out for specified delinquent subtypes; (b) broaden the range and variety of offenders to whom differential treatment may be usefully applied; (c) continue to isolate the factors which may be responsible for, and/or essential to, the implementation and success of the differential treatment approaches which have been developed to date; and (d) continue refining and expanding the Differential Treatment Model. The Settings For Differential Treatment of Delinquents experiment would attempt to determine whether specified treatment objectives with certain youthful offenders could be better accomplished in the event that treatment were to begin, not within the community proper, but within a differential treatment-oriented, residential (Regional) treatment setting. To carry out this five year experiment, approximately 460 offenders would be classified and then randomly assigned to one of four experimental conditions. Several behavioral as well as attitudinal indices will be utilized as dependent variables. Univariate and multivariate analyses will be carried out relative to these measures, and in relation to data which bears upon the treatment process as well.